Abstract for: Narratives of stewardship and resistance: Deciphering divergent sustainability decision-making among livestock farmers
Effective environmental stewardship among livestock farmers and herders is shaped by interconnected socio-economic, ideological and environmental factors. Understanding the feedback mechanisms behind sustainability decision-making is crucial for promoting agri-food sustainability. This study applies AI-enhanced graph-based analytics and system dynamics modeling to examine how beliefs, values, and narratives influence farmers' behaviors amid policy intervention, market competition, and evolving environmental and social landscapes. With a knowledge graph constructed using Neo4j and natural language processing (NLP), this study analyzes texts from social media, literature, and other sources to identify causal feedback loops in sustainability discourse. Causal loop diagrams (CLDs) are constructed to map two broad orientations: one emphasizing adaptive stewardship and long-term ecological resilience, and another prioritizing traditional place-based identities and continuity of established practices. These structures provide a foundation for quantitative system dynamics modeling. The current findings reveal key feedback loops shaping agri-food sustainability narratives. Some mechanisms reinforce proactive environmental strategies, while others contribute to resistance or inertia in adopting new management practices. Using the knowledge graph, this work-in-progress study will extract stock (e.g., land health, resource availability) and flow variables (e.g., policy shifts, knowledge diffusion) to develop a system dynamics model, with structured causal relationships simulated in Stella for policy and sustainability scenario analysis. Rather than making normative judgments, this study provides insights into how environmental and sustainability narratives emerge, persist, and evolve over time. By recognizing the socio-economic, ideological and environmental contexts influencing decision-making, this work-in-progress research will identify leverage points for shaping sustainable agri-food outcomes. The approach used contributes to a deeper understanding of social-environmental systems, offering quantitative insights for policy in agri-food sustainability. *AI was used in this research for enhancement of modeling stages, and for automation and efficiency.